CODEWITHSUNDEEP
algorithm
A. Robben
A. Robben

Reputation: 27

What is the purpose of this method?

in an interview question, I got asked the following:

What is the purpose of the below method, and how can we rewrite it?

public int question_1(int a, int b)
{
    while (a > b)
    {
        a -= b;
    }
    return a;
}

at first I thought it is equivalent to a%b, but it is not since it is "while (a > b)" and not "while ( a >= b)".

Thanks

Upvotes: 0

Views: 325

Answers (2)

Matt Timmermans
Matt Timmermans

Reputation: 59154

I would guess that its purpose is to compute a%b for positive ints, and that it has a bug.

If I saw this in production, I would have to check the uses of this function to see if question_1(n,n) == n is really correct. If so, I'd add a comment indicating why that is so. Otherwise I'd fix it.

It either case, it could be rewritten to use the % operator instead of a loop. If it's correct, it could be rewritten like this:

public int question_1(int a, int b)
{
    if (a>b)
    {
        a = ((a-1)%b) + 1;
    }
    return a;
}

This is not the same in its handling of negative numbers, though, so again you'd have to check to make sure that's OK.

The reason I provide this answer when @ruakh has already provided such a carefully considered answer is that this is an interview question, so it's best if you take the opportunity to show how you would approach a problem like this on the job.

You don't really want to give the impression that you would spend a long time and a lot of effort thinking carefully about such a simple problem -- if you have to spend that much effort to solve a simple problem, imagine what you would spend on a big one!

At the same time, you want to demonstrate that you recognize the possible bug, and take the initiative to fix it or to spare future engineers the same task.

Upvotes: 1

ruakh
ruakh

Reputation: 183270

Honestly, it's impossible to know the purpose of a method just by reading its implementation, even if we assume that it's bug-free.

But, we can start by documenting its behaviors:

  • If b is positive:
    • If a is positive, the method returns the least positive integer that is congruent to a modulo b. (For example, given 15 and 10, it will return 5; given 30 and 10, it will return 10.)
    • Otherwise, the method returns a.
  • If b is zero:
    • If a is positive, the method loops forever.
    • Otherwise, the method returns a.
  • If b is negative:
    • If a ≤ b, the method returns a.
    • Otherwise, the method's behavior depends on the language, since it will increase a until it's no longer greater than b. If the language defines integer arithmetic using "wraparound" rules, then the method will loop for a long time, then finally return a very negative number (unless b is itself very negative, in which case, depending on the value of a, the function might loop forever).

and given these, we can infer that the behaviors with zero and negative numbers are bizarre enough that the method is probably actually only intended to be used with positive numbers. So its behavior can be summarized as:

  • If a and b are both positive, then the method returns the least positive integer that is congruent to a modulo b.

If the above inference is correct, then the method can be rewritten as:

public int question_1(int a, int b) {
    if (a <= 0 || b <= 0)
        throw new IllegalArgumentException();
    return (a - 1) % b + 1;
}

Upvotes: 4

Related Questions